ddia
Skill17 starsChanged 9 months agoMITView on GitHub
Reference for distributed systems and data architecture concepts from Martin Kleppmann's "Designing Data-Intensive Applications." Use this skill when discussing: - Database selection and data modeling - Replication and high availability - Partitioning/sharding strategies - Distributed transactions - Consistency models and guarantees - Stream vs batch processing - Event sourcing and CQRS "In the presence of a network partition, choose Consistency OR Availability."
In this repository7 other files it gives its agents.
- clean-codeSkill
- clrs-algorithmsSkill
- dddSkill
- gof-patternsSkill
- pragmatic-programmerSkill
- sicpSkill
- software-patternsSkill
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 7
Use this skill when discussing:
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Designing Data-Intensive Applications Skill
Reference for distributed systems and data architecture concepts from Martin Kleppmann's "Designing Data-Intensive Applications."
Activation Triggers
Use this skill when discussing:
- Database selection and data modeling
- Replication and high availability
- Partitioning/sharding strategies
- Distributed transactions
- Consistency models and guarantees
- Stream vs batch processing
- Event sourcing and CQRS
Quick Reference
Data Models
| Model | Best For | Trade-offs |
| Relational | Complex queries, joins, ACID | Schema rigidity, scaling writes |
| Document | Hierarchical data, flexibility | Poor joins, denormalization |
| Graph | Highly connected data | Specialized queries, complexity |
| Wide-Column | Time series, analytics | Limited query patterns |
Storage Engines
| Engine | Optimized For | Examples |
| B-Tree | Read-heavy, random access | PostgreSQL, MySQL |
| LSM-Tree | Write-heavy, sequential | Cassandra, RocksDB, LevelDB |
| Column Store | Analytics, aggregations | ClickHouse, Parquet |
Replication Strategies
| Strategy | Consistency | Availability | Use Case |
| Single Leader | Strong | Medium | Traditional RDBMS |
| Multi-Leader | Eventual | High | Multi-datacenter |
| Leaderless | Eventual | Highest | High availability |
Partitioning Strategies
| Strategy | Description | Pros | Cons |
| Range | Partition by key ranges | Efficient range queries | Hot spots |
| Hash | Partition by hash of key | Even distribution | No range queries |
| Composite | Combine range + hash | Balanced | Complexity |
Consistency Models
| Model | Guarantee | Performance |
| Linearizable | Strongest (appears sequential) | Slowest |
| Sequential | Operations ordered per client | Medium |
| Causal | Cause-effect preserved | Good |
| Eventual | Will converge eventually | Fastest |
Transaction Isolation Levels
| Level | Dirty Read | Non-Repeatable | Phantom |
| Read Uncommitted | ✗ | ✗ | ✗ |
| Read Committed | ✓ | ✗ | ✗ |
| Repeatable Read | ✓ | ✓ | ✗ |
| Serializable | ✓ | ✓ | ✓ |
CAP Theorem
"In the presence of a network partition, choose Consistency OR Availability."
| Choice | Behavior | Examples |
| CP | Reject requests if can't guarantee consistency | ZooKeeper, HBase |
| AP | Accept requests, allow inconsistency | Cassandra, DynamoDB |
Batch vs Stream Processing
| Aspect | Batch | Stream |
| Latency | High (hours/days) | Low (seconds/minutes) |
| Data | Bounded, complete | Unbounded, continuous |
| Processing | MapReduce, Spark | Kafka, Flink, Storm |
| Use Case | Analytics, ETL | Real-time alerts, dashboards |
Directory Structure
ddia/
├── SKILL.md
├── data-models/
│ ├── relational.md
│ ├── document.md
│ └── graph.md
├── storage/
│ ├── b-trees.md
│ ├── lsm-trees.md
│ └── column-storage.md
├── replication/
│ ├── leader-follower.md
│ ├── multi-leader.md
│ └── leaderless.md
├── partitioning/
│ ├── strategies.md
│ └── rebalancing.md
├── transactions/
│ ├── acid.md
│ ├── isolation-levels.md
│ └── distributed-transactions.md
├── consistency/
│ ├── models.md
│ └── linearizability.md
├── consensus/
│ └── algorithms.md
└── processing/
├── batch.md
├── stream.md
└── event-sourcing.mdUsage Examples
Choosing a Database
Question: "Should I use PostgreSQL or MongoDB?"
Consider:
- Data relationships → See data-models/
- Query patterns → See storage/
- Scale requirements → See partitioning/
- Consistency needs → See consistency/Designing for Scale
Question: "How do I handle millions of users?"
Consider:
- Read scaling → See replication/leader-follower.md
- Write scaling → See partitioning/strategies.md
- Geographic distribution → See replication/multi-leader.mdHandling Failures
Question: "What happens when a node fails?"
Consider:
- Data durability → See replication/
- Consistency trade-offs → See consistency/models.md
- Recovery → See consensus/algorithms.mdBased on concepts from "Designing Data-Intensive Applications" by Martin Kleppmann.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.